Merge branches 'clk-socfpga', 'clk-doc', 'clk-qcom', 'clk-vc5' and 'clk-bcm' into clk-next

 - Enable CPU clks on Qualcomm IPQ6018 SoCs
 - Enable CPU clks on Qualcomm MSM8996 SoCs
 - GPU clk support for Qualcomm SM8150 and SM8250 SoCs
 - Audio clks on Qualcomm SC7180 SoCs
 - Make defines for bcm63xx-gate clks to use in DT
 - Support gate clks on BCM6318 SoCs
 - Add HDMI clks for BCM2711 SoCs
 - Support BCM2711 SoC firmware clks

* clk-socfpga:
  clk: socfpga: agilex: mpu_l2ram_clk should be mpu_ccu_clk
  clk: socfpga: agilex: add nand_x_clk and nand_ecc_clk
  dt-bindings: agilex: add NAND_X_CLK and NAND_ECC_CLK

* clk-doc:
  clk: Clean up kernel-doc errors
  clk: <linux/clk-provider.h>: drop a duplicated word
  clk: add function documentation for clk_hw_round_rate()

* clk-qcom: (38 commits)
  dt-bindings: clock: Fix YAML schemas for LPASS clocks on SC7180
  clk: qcom: gcc-sdm660: Fix up gcc_mss_mnoc_bimc_axi_clk
  clk: qcom: gcc-sdm660: Add missing modem reset
  clk: qcom: lpass: Add support for LPASS clock controller for SC7180
  clk: qcom: gcc: Add support for GCC LPASS clock for SC7180
  dt-bindings: clock: Add YAML schemas for LPASS clocks on SC7180
  clk: qcom: gdsc: Add support to enable retention of GSDCR
  clk: qcom: Export gdsc_gx_do_nothing_enable() to modules
  clk: qcom: Add graphics clock controller driver for SM8250
  clk: qcom: Add graphics clock controller driver for SM8150
  clk: qcom: add common gdsc_gx_do_nothing_enable for gpucc drivers
  dt-bindings: clock: add SM8250 QCOM Graphics clock bindings
  dt-bindings: clock: add SM8150 QCOM Graphics clock bindings
  dt-bindings: clock: combine qcom,sdm845-gpucc and qcom,sc7180-gpucc
  clk: qcom: gcc: remove unnecessary vco_table from SM8150
  clk: qcom: clk-alpha-pll: use the right PCAL_DONE value for lucid pll
  clk: qcom: clk-alpha-pll: same regs and ops for trion and lucid
  clk: qcom: clk-alpha-pll: remove unused/incorrect PLL_CAL_VAL
  clk: qcom: gcc: fix sm8150 GPU and NPU clocks
  dt-bindings: clock: Fix qcom,msm8996-apcc yaml syntax
  ...

* clk-vc5:
  clk: vc5: use a dedicated struct to describe the output drivers
  dt-bindings: clk: versaclock5: convert to yaml
  MAINTAINERS: take over IDT VersaClock 5 clock driver
  dt-bindings: clk: versaclock5: fix 'idt' prefix typos
  clk: vc5: Add memory check to prevent oops
  clk: vc5: fix use of memory after it has been kfree'd
  clk: vc5: Enable addition output configurations of the Versaclock
  dt: Add additional option bindings for IDT VersaClock
  clk: vc5: Allow Versaclock driver to support multiple instances

* clk-bcm: (44 commits)
  clk: bcm2835: Do not use prediv with bcm2711's PLLs
  dt-bindings: arm: bcm: Add a select to the RPI Firmware binding
  clk: bcm: dvp: Add missing module informations
  clk: bcm: rpi: Remove the quirks for the CPU clock
  clk: bcm2835: Don't cache the PLLB rate
  clk: bcm2835: Allow custom CCF flags for the PLLs
  Revert "clk: bcm2835: remove pllb"
  clk: bcm: rpi: Give firmware clocks a name
  clk: bcm: rpi: Discover the firmware clocks
  clk: bcm: rpi: Add an enum for the firmware clocks
  clk: bcm: rpi: Add DT provider for the clocks
  clk: bcm: rpi: Make the PLLB registration function return a clk_hw
  clk: bcm: rpi: Split pllb clock hooks
  clk: bcm: rpi: Rename is_prepared function
  clk: bcm: rpi: Pass the clocks data to the firmware function
  clk: bcm: rpi: Add clock id to data
  clk: bcm: rpi: Create a data structure for the clocks
  clk: bcm: rpi: Use CCF boundaries instead of rolling our own
  clk: bcm: rpi: Make sure the clkdev lookup is removed
  clk: bcm: rpi: Switch to clk_hw_register_clkdev
  ...

clk: bcm2835: Do not use prediv with bcm2711's PLLs

Contrary to previous SoCs, bcm2711 doesn't have a prescaler in the PLL
feedback loop. Bypass it by zeroing fb_prediv_mask when running on
bcm2711.

Note that, since the prediv configuration bits were re-purposed, this
was triggering miscalculations on all clocks hanging from the VPU clock,
notably the aux UART, making its output unintelligible.

Fixes: 42de9ad400af ("clk: bcm2835: Add BCM2711_CLOCK_EMMC2 support")
Reported-by: Nathan Chancellor <natechancellor@gmail.com>
Signed-off-by: Nicolas Saenz Julienne <nsaenzjulienne@suse.de>
Link: https://lore.kernel.org/r/20200730182619.23246-1-nsaenzjulienne@suse.de
Tested-by: Nathan Chancellor <natechancellor@gmail.com>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

dt-bindings: clock: Fix YAML schemas for LPASS clocks on SC7180

The YAML schemas that landed forgot one clock: "bi_tcxo".  Presumably
the bindings were developed against the v4 version of the driver and
when the ".name" was removed in v5 of the driver things broke.

While touching this, add the needed includes in each example.  I
believe both examples are supposed to be independent of each other.

Let's fix the bindings.

Fixes: 381cc6f97cda ("dt-bindings: clock: Add YAML schemas for LPASS clocks on SC7180")
Signed-off-by: Douglas Anderson <dianders@chromium.org>
Link: https://lore.kernel.org/r/20200731133006.1.Iee81b115f5be50d6d69500fe1bda11bba6e16143@changeid
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

clk: qcom: gcc-sdm660: Fix up gcc_mss_mnoc_bimc_axi_clk

Add missing halt_check, hwcg_reg and hwcg_bit properties.
These were likely omitted when porting the driver upstream.

Signed-off-by: Konrad Dybcio <konradybcio@gmail.com>
Link: https://lore.kernel.org/r/20200726111215.22361-9-konradybcio@gmail.com
Fixes: f2a76a2955c0 ("clk: qcom: Add Global Clock controller (GCC) driver for SDM660")
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

clk: qcom: gcc-sdm660: Add missing modem reset

This will be required in order to support the
modem upstream.

Signed-off-by: Konrad Dybcio <konradybcio@gmail.com>
Link: https://lore.kernel.org/r/20200726111215.22361-2-konradybcio@gmail.com
Fixes: f2a76a2955c0 ("clk: qcom: Add Global Clock controller (GCC) driver for SDM660")
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

clk: qcom: lpass: Add support for LPASS clock controller for SC7180

The Low Power Audio subsystem clocks are required for Audio client
to be able to request for the clocks and power domains.

Signed-off-by: Taniya Das <tdas@codeaurora.org>
Link: https://lore.kernel.org/r/1595606878-2664-5-git-send-email-tdas@codeaurora.org
[sboyd@kernel.org: Drop unused ret in probe function]
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

clk: qcom: gcc: Add support for GCC LPASS clock for SC7180

Add the GCC lpass clock which is required to access the LPASS core
clocks.

Signed-off-by: Taniya Das <tdas@codeaurora.org>
Link: https://lore.kernel.org/r/1595606878-2664-4-git-send-email-tdas@codeaurora.org
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

dt-bindings: clock: Add YAML schemas for LPASS clocks on SC7180

The LPASS(Low Power Audio Subsystem) clock provider have a bunch of generic
properties that are needed in a device tree. Also add clock ids for GCC
LPASS and LPASS Core clock IDs for LPASS client to request for the clocks.

Signed-off-by: Taniya Das <tdas@codeaurora.org>
Reviewed-by: Rob Herring <robh@kernel.org>
Link: https://lore.kernel.org/r/1595606878-2664-3-git-send-email-tdas@codeaurora.org
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

clk: qcom: gdsc: Add support to enable retention of GSDCR

Add support for the RETAIN_FF_ENABLE feature which enables the
usage of retention registers. These registers maintain their
state after disabling and re-enabling a GDSC.

Signed-off-by: Taniya Das <tdas@codeaurora.org>
Link: https://lore.kernel.org/r/1595606878-2664-2-git-send-email-tdas@codeaurora.org
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

clk: Clean up kernel-doc errors

Two things aren't documented causing kernel-doc to fail when checking
the core clk.c file. Fix them so that this file is clean.

Signed-off-by: Stephen Boyd <sboyd@kernel.org>
Link: https://lore.kernel.org/r/20200622090935.213833-1-sboyd@kernel.org

clk: qcom: Export gdsc_gx_do_nothing_enable() to modules

A clk driver can be a module but the gdsc code is in the common module.
Export this symbol so that allmodconfig builds keep working.

Cc: Jonathan Marek <jonathan@marek.ca>
Fixes: 0638226dd095 ("clk: qcom: add common gdsc_gx_do_nothing_enable for gpucc drivers")
Signed-off-by: Stephen Boyd <sboyd@kernel.org>
Link: https://lore.kernel.org/r/20200724094025.3261266-1-sboyd@kernel.org

clk: qcom: Add graphics clock controller driver for SM8250

Add support for the graphics clock controller found on SM8250
based devices.

This is initially copied from the downstream kernel, but has
been modified to more closely match the upstream sc7180 driver.

Signed-off-by: Jonathan Marek <jonathan@marek.ca>
Tested-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Link: https://lore.kernel.org/r/20200709135251.643-12-jonathan@marek.ca
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

clk: qcom: Add graphics clock controller driver for SM8150

Add support for the graphics clock controller found on SM8150
based devices.

This is initially copied from the downstream kernel, but has
been modified to more closely match the upstream sc7180 driver.

Signed-off-by: Jonathan Marek <jonathan@marek.ca>
Tested-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Link: https://lore.kernel.org/r/20200709135251.643-11-jonathan@marek.ca
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

clk: qcom: add common gdsc_gx_do_nothing_enable for gpucc drivers

All gpucc drivers need this, so move it to common code instead of
duplicating it in every gpucc driver.

Signed-off-by: Jonathan Marek <jonathan@marek.ca>
Tested-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Link: https://lore.kernel.org/r/20200709135251.643-10-jonathan@marek.ca
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

dt-bindings: clock: add SM8250 QCOM Graphics clock bindings

Add device tree bindings for graphics clock controller for
Qualcomm Technology Inc's SM8250 SoCs.

Signed-off-by: Jonathan Marek <jonathan@marek.ca>
Tested-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Link: https://lore.kernel.org/r/20200709135251.643-9-jonathan@marek.ca
Reviewed-by: Rob Herring <robh@kernel.org>
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

dt-bindings: clock: add SM8150 QCOM Graphics clock bindings

Add device tree bindings for graphics clock controller for
Qualcomm Technology Inc's SM8150 SoCs.

Signed-off-by: Jonathan Marek <jonathan@marek.ca>
Tested-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Link: https://lore.kernel.org/r/20200709135251.643-8-jonathan@marek.ca
Reviewed-by: Rob Herring <robh@kernel.org>
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

dt-bindings: clock: combine qcom,sdm845-gpucc and qcom,sc7180-gpucc

These two bindings are almost identical, so combine them into one. This
will make it easier to add the sm8150 and sm8250 gpucc bindings.

Signed-off-by: Jonathan Marek <jonathan@marek.ca>
Tested-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Link: https://lore.kernel.org/r/20200709135251.643-7-jonathan@marek.ca
Reviewed-by: Rob Herring <robh@kernel.org>
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

clk: qcom: gcc: remove unnecessary vco_table from SM8150

The fixed alpha pll ops only use it for clamping in round_rate, which is
unnecessary. This is consistent with SM8250 GCC not using vco_table.

Signed-off-by: Jonathan Marek <jonathan@marek.ca>
Tested-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Link: https://lore.kernel.org/r/20200709135251.643-6-jonathan@marek.ca
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

clk: qcom: clk-alpha-pll: use the right PCAL_DONE value for lucid pll

Lucid PCAL_DONE is different from trion.

Signed-off-by: Jonathan Marek <jonathan@marek.ca>
Tested-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Link: https://lore.kernel.org/r/20200709135251.643-5-jonathan@marek.ca
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

clk: qcom: clk-alpha-pll: same regs and ops for trion and lucid

Fixed ops were already identical, this adds support for non-fixed ops by
sharing between trion and lucid.

This also changes the names for trion ops to be consistent with the rest.

Note LUCID_PCAL_DONE is renamed to TRION_PCAL_DONE because it is wrong for
lucid, LUCID_PCAL_DONE should be BIT(27). Next patch will address this.

Signed-off-by: Jonathan Marek <jonathan@marek.ca>
Tested-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Link: https://lore.kernel.org/r/20200709135251.643-4-jonathan@marek.ca
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

clk: qcom: clk-alpha-pll: remove unused/incorrect PLL_CAL_VAL

0x44 isn't a register offset, it is the value that goes into CAL_L_VAL.

Fixes: 548a909597d5 ("clk: qcom: clk-alpha-pll: Add support for Trion PLLs")
Signed-off-by: Jonathan Marek <jonathan@marek.ca>
Tested-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Link: https://lore.kernel.org/r/20200709135251.643-3-jonathan@marek.ca
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

clk: qcom: gcc: fix sm8150 GPU and NPU clocks

Fix the parents and set BRANCH_HALT_SKIP. From the downstream driver it
should be a 500us delay and not skip, however this matches what was done
for other clocks that had 500us delay in downstream.

Fixes: f73a4230d5bb ("clk: qcom: gcc: Add GPU and NPU clocks for SM8150")
Signed-off-by: Jonathan Marek <jonathan@marek.ca>
Tested-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
Link: https://lore.kernel.org/r/20200709135251.643-2-jonathan@marek.ca
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

clk: vc5: use a dedicated struct to describe the output drivers

Reusing the generic struct vc5_hw_data for all blocks is handy. However it
implies we allocate space the div_int and div_frc fields even for the
output drivers where they are unused, and the clk_output_cfg0 and
clk_output_cfg0_mask fields for all components even though they are used
only for the output drivers.

Use a dedicated struct for the output drivers so that each block uses
exactly the fields it needs, not more.

Signed-off-by: Luca Ceresoli <luca@lucaceresoli.net>
Link: https://lore.kernel.org/r/20200723072603.1795-1-luca@lucaceresoli.net
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

dt-bindings: clk: versaclock5: convert to yaml

Convert to yaml the VersaClock bindings document. The mapping between
clock specifier and physical pins cannot be described formally in yaml
schema, then keep it verbatim in the description field.

Signed-off-by: Luca Ceresoli <luca@lucaceresoli.net>
Link: https://lore.kernel.org/r/20200723074112.3159-4-luca@lucaceresoli.net
Reviewed-by: Rob Herring <robh@kernel.org>
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

MAINTAINERS: take over IDT VersaClock 5 clock driver

Marek has been the primary developer of this driver (thanks!). Now as
he is not working on it anymore he suggested I take over maintainership.

Cc: Marek Vasut <marek.vasut@gmail.com>
Signed-off-by: Luca Ceresoli <luca@lucaceresoli.net>
Link: https://lore.kernel.org/r/20200723074112.3159-3-luca@lucaceresoli.net
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

dt-bindings: clk: versaclock5: fix 'idt' prefix typos

'idt' is misspelled 'itd' in a few places, fix it.

Fixes: 34662f6e3084 ("dt: Add additional option bindings for IDT VersaClock")
Signed-off-by: Luca Ceresoli <luca@lucaceresoli.net>
Reviewed-by: Rob Herring <robh@kernel.org>
Link: https://lore.kernel.org/r/20200723074112.3159-2-luca@lucaceresoli.net
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

dt-bindings: arm: bcm: Add a select to the RPI Firmware binding

The RaspberryPi firmware binding uses two compatible, include simple-bus.
The select statement generated by default will thus select any node that
has simple-bus, not all of them being the raspberrypi firmware node.

This results in warnings being wrongfully reported. Let's add a custom
select statement to fix that.

Fixes: d4c708c032df ("dt-bindings: arm: bcm: Convert BCM2835 firmware binding to YAML")
Signed-off-by: Maxime Ripard <maxime@cerno.tech>
Link: https://lore.kernel.org/r/20200626115433.125735-1-maxime@cerno.tech
Acked-by: Florian Fainelli <f.fainelli@gmail.com>
Acked-by: Rob Herring <robh@kernel.org>
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

clk: vc5: Add memory check to prevent oops

When getting the names of the child nodes, kasprintf is used to
allocate memory which is used to create the string for the node
name.  Unfortunately, there is no memory check to determine
if this allocation fails, it may cause an error when trying
to get child node name.

This patch will check if the memory allocation fails, and returns
and -ENOMEM error instead of blindly moving on.

Fixes: 260249f929e8 ("clk: vc5: Enable addition output configurations of the Versaclock")
Suggested-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Adam Ford <aford173@gmail.com>
Reviewed-by: Luca Ceresoli <luca@lucaceresoli.net>
Link: https://lore.kernel.org/r/20200716122620.4538-1-aford173@gmail.com
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

clk: vc5: fix use of memory after it has been kfree'd

There are a several places where printing an error message of
init.name occurs after init.name has been kfree'd. Also the failure
message is duplicated each time in the code. Fix this by adding
a registration error failure path for these cases, moving the
duplicated error messages to one common point and kfree'ing init.name
only after it has been used.

Changes also shrink the object code size by 171 bytes (x86-64, gcc 9.3):

Before:
   text    data     bss     dec     hex filename
  21057    3960      64   25081    61f9 drivers/clk/clk-versaclock5.o

After:
   text    data     bss     dec     hex filename
  20886    3960      64   24910    614e drivers/clk/clk-versaclock5.o

Addresses-Coverity: ("Use after free")
Fixes: f491276a5168 ("clk: vc5: Allow Versaclock driver to support multiple instances")
Signed-off-by: Colin Ian King <colin.king@canonical.com>
Link: https://lore.kernel.org/r/20200625132736.88832-1-colin.king@canonical.com
Reviewed-by: Luca Ceresoli <luca@lucaceresoli.net>
[sboyd@kernel.org: Drop stray newline]
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

dt-bindings: clock: Fix qcom,msm8996-apcc yaml syntax

Fix errors reported by dt_binding_check.
- Fix literal block scalar for dts example
- Fix schema identifier URI

Signed-off-by: Loic Poulain <loic.poulain@linaro.org>
Link: https://lore.kernel.org/r/1595326714-20485-1-git-send-email-loic.poulain@linaro.org
Reviewed-by: Rob Herring <robh@kernel.org>
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

clk: qcom: gcc: Make disp gpll0 branch aon for sc7180/sdm845

The display gpll0 branch clock inside GCC needs to always be enabled.
Otherwise the AHB clk (disp_cc_mdss_ahb_clk_src) for the display clk
controller (dispcc) will stop clocking while sourcing from gpll0 when
this branch inside GCC is turned off during unused clk disabling. We can
never turn this branch off because the AHB clk for the display subsystem
is needed to read/write any registers inside the display subsystem
including clk related ones. This makes this branch a really easy way to
turn off AHB access to the display subsystem and cause all sorts of
mayhem. Let's just make the clk ops keep the clk enabled forever and
ignore any attempts to disable this clk so that dispcc accesses keep
working.

Signed-off-by: Taniya Das <tdas@codeaurora.org>
Reported-by: Evan Green <evgreen@chromium.org>
Link: https://lore.kernel.org/r/1594796050-14511-1-git-send-email-tdas@codeaurora.org
Fixes: 17269568f726 ("clk: qcom: Add Global Clock controller (GCC) driver for SC7180")
Fixes: 06391eddb60a ("clk: qcom: Add Global Clock controller (GCC) driver for SDM845")
[sboyd@kernel.org: Fill out commit text more]
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

ipq806x: gcc: add support for child probe

Add support for child probing needed for tsens driver that share the
same regs of gcc for this platform.

Signed-off-by: Ansuel Smith <ansuelsmth@gmail.com>
Reviewed-by: Amit Kucheria <amit.kucheria@linaro.org>
Link: https://lore.kernel.org/r/20200716022817.30439-2-ansuelsmth@gmail.com
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

clk: qcom: msm8996: Make symbol 'cpu_msm8996_clks' static

The sparse tool complains as follows:

drivers/clk/qcom/clk-cpu-8996.c:341:19: warning:
 symbol 'cpu_msm8996_clks' was not declared. Should it be static?

This variable is not used outside of clk-cpu-8996.c, so this commit
marks it static.

Fixes: 03e342dc45c9 ("clk: qcom: Add CPU clock driver for msm8996")
Reported-by: Hulk Robot <hulkci@huawei.com>
Signed-off-by: Wei Yongjun <weiyongjun1@huawei.com>
Link: https://lore.kernel.org/r/20200714142155.35085-1-weiyongjun1@huawei.com
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

clk: qcom: ipq8074: Add correct index for PCIe clocks

The PCIe clocks GCC_PCIE0_AXI_S_BRIDGE_CLK, GCC_PCIE0_RCHNG_CLK_SRC,
GCC_PCIE0_RCHNG_CLK are wrongly added to the gcc reset group.

Move them to the gcc clock group.

Reported-by: kernel test robot <lkp@intel.com>
Signed-off-by: Sivaprakash Murugesan <sivaprak@codeaurora.org>
Link: https://lore.kernel.org/r/1594877570-9280-1-git-send-email-sivaprak@codeaurora.org
Fixes: e7fb524cfcca ("dt-bindings: clock: qcom: ipq8074: Add missing bindings for PCIe")
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

clk: <linux/clk-provider.h>: drop a duplicated word

Drop the repeated word "not" in a comment.

Signed-off-by: Randy Dunlap <rdunlap@infradead.org>
Cc: Michael Turquette <mturquette@baylibre.com>
Cc: Stephen Boyd <sboyd@kernel.org>
Cc: linux-clk@vger.kernel.org
Link: https://lore.kernel.org/r/20200719002830.20319-1-rdunlap@infradead.org
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

clk: qcom: smd: Add support for MSM8992/4 rpm clocks

Add rpm smd clocks, PMIC and bus clocks which are required on MSM8992,
MSM8994 (and APQ variants) for clients to vote on.

Signed-off-by: Konrad Dybcio <konradybcio@gmail.com>
Link: https://lore.kernel.org/r/20200623230018.303776-1-konradybcio@gmail.com
[sboyd@kernel.org: Fixed up binding numbers]
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

clk: qcom: ipq8074: Add missing clocks for pcie

Add missing clocks and resets for pcie port0 of ipq8074 devices.

Co-developed-by: Selvam Sathappan Periakaruppan <speriaka@codeaurora.org>
Signed-off-by: Selvam Sathappan Periakaruppan <speriaka@codeaurora.org>
Signed-off-by: Sivaprakash Murugesan <sivaprak@codeaurora.org>
Link: https://lore.kernel.org/r/1593940680-2363-5-git-send-email-sivaprak@codeaurora.org
[sboyd@kernel.org: Make freq table static const]
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

dt-bindings: clock: qcom: ipq8074: Add missing bindings for PCIe

Add missing clock bindings for PCIe port0 of ipq8074.

Co-developed-by: Selvam Sathappan Periakaruppan <speriaka@codeaurora.org>
Signed-off-by: Selvam Sathappan Periakaruppan <speriaka@codeaurora.org>
Signed-off-by: Sivaprakash Murugesan <sivaprak@codeaurora.org>
Link: https://lore.kernel.org/r/1593940680-2363-4-git-send-email-sivaprak@codeaurora.org
[sboyd@kernel.org: Clean up commit text subject]
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

clk: qcom: Add CPU clock driver for msm8996

Each of the CPU clusters (Power and Perf) on msm8996 are
clocked via 2 PLLs, a primary and alternate. There are also
2 Mux'es, a primary and secondary all connected together
as shown below

                             +-------+
              XO             |       |
          +------------------>0      |
                             |       |
                   PLL/2     | SMUX  +----+
                     +------->1      |    |
                     |       |       |    |
                     |       +-------+    |    +-------+
                     |                    +---->0      |
                     |                         |       |
+---------------+    |             +----------->1      | CPU clk
|Primary PLL    +----+ PLL_EARLY   |           |       +------>
|               +------+-----------+    +------>2 PMUX |
+---------------+      |                |      |       |
                       |   +------+     |   +-->3      |
                       +--^+  ACD +-----+   |  +-------+
+---------------+          +------+         |
|Alt PLL        |                           |
|               +---------------------------+
+---------------+         PLL_EARLY

The primary PLL is what drives the CPU clk, except for times
when we are reprogramming the PLL itself (for rate changes) when
we temporarily switch to an alternate PLL. A subsequent patch adds
support to switch between primary and alternate PLL during rate
changes.

The primary PLL operates on a single VCO range, between 600MHz
and 3GHz. However the CPUs do support OPPs with frequencies
between 300MHz and 600MHz. In order to support running the CPUs
at those frequencies we end up having to lock the PLL at twice
the rate and drive the CPU clk via the PLL/2 output and SMUX.

So for frequencies above 600MHz we follow the following path
 Primary PLL --> PLL_EARLY --> PMUX(1) --> CPU clk
and for frequencies between 300MHz and 600MHz we follow
 Primary PLL --> PLL/2 --> SMUX(1) --> PMUX(0) --> CPU clk

ACD stands for Adaptive Clock Distribution and is used to
detect voltage droops.

Signed-off-by: Rajendra Nayak <rnayak@codeaurora.org>
Rajendra Nayak: Initial RFC - https://lkml.org/lkml/2016/9/29/84

Signed-off-by: Ilia Lin <ilialin@codeaurora.org>
Ilia Lin:  - reworked clock registering
           - Added clock-tree diagram
           - non-builtin support
           - clock notifier on rate change
           - https://lkml.org/lkml/2018/5/24/123

Signed-off-by: Loic Poulain <loic.poulain@linaro.org>
Loic Poulain: - fixed driver remove / clk deregistering
              - Removed useless memory barriers
              - devm usage when possible
              - Fixed Kconfig depends

Link: https://lore.kernel.org/r/1593766185-16346-3-git-send-email-loic.poulain@linaro.org
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

dt-bindings: clk: qcom: Add bindings for CPU clock for msm8996

Each of the CPU clusters (Power and Perf) on msm8996 are
clocked via 2 PLLs, a primary and alternate. There are also
2 Mux'es, a primary and secondary all connected together
as shown below

                             +-------+
              XO             |       |
          +------------------>0      |
                             |       |
                   PLL/2     | SMUX  +----+
                     +------->1      |    |
                     |       |       |    |
                     |       +-------+    |    +-------+
                     |                    +---->0      |
                     |                         |       |
+---------------+    |             +----------->1      | CPU clk
|Primary PLL    +----+ PLL_EARLY   |           |       +------>
|               +------+-----------+    +------>2 PMUX |
+---------------+      |                |      |       |
                       |   +------+     |   +-->3      |
                       +--^+  ACD +-----+   |  +-------+
+---------------+          +------+         |
|Alt PLL        |                           |
|               +---------------------------+
+---------------+         PLL_EARLY

The primary PLL is what drives the CPU clk, except for times
when we are reprogramming the PLL itself (for rate changes) when
we temporarily switch to an alternate PLL. A subsequent patch adds
support to switch between primary and alternate PLL during rate
changes.

The primary PLL operates on a single VCO range, between 600MHz
and 3GHz. However the CPUs do support OPPs with frequencies
between 300MHz and 600MHz. In order to support running the CPUs
at those frequencies we end up having to lock the PLL at twice
the rate and drive the CPU clk via the PLL/2 output and SMUX.

Signed-off-by: Ilia Lin <ilialin@codeaurora.org>
Reviewed-by: Rob Herring <robh@kernel.org>
Link: https://lore.kernel.org/r/1593766185-16346-4-git-send-email-loic.poulain@linaro.org
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

soc: qcom: Separate kryo l2 accessors from PMU driver

The driver provides kernel level API for other drivers
to access the MSM8996 L2 cache registers.
Separating the L2 access code from the PMU driver and
making it public to allow other drivers use it.
The accesses must be separated with a single spinlock,
maintained in this driver.

Signed-off-by: Ilia Lin <ilialin@codeaurora.org>
Signed-off-by: Loic Poulain <loic.poulain@linaro.org>
Link: https://lore.kernel.org/r/1593766185-16346-2-git-send-email-loic.poulain@linaro.org
Acked-by: Will Deacon <will@kernel.org>
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

clk: qcom: Fix return value check in apss_ipq6018_probe()

In case of error, the function dev_get_regmap() returns NULL pointer
not ERR_PTR(). The IS_ERR() test in the return value check should be
replaced with NULL test.

Fixes: 5e77b4ef1b19 ("clk: qcom: Add ipq6018 apss clock controller")
Reported-by: Hulk Robot <hulkci@huawei.com>
Signed-off-by: Wei Yongjun <weiyongjun1@huawei.com>
Link: https://lore.kernel.org/r/20200628023055.50608-1-weiyongjun1@huawei.com
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

clk: bcm: dvp: Add missing module informations

The driver for the DVP controller in the BCM2711 was missing the MODULE_*
macros resulting in a modpost warning at compilation.

Fixes: 1bc95972715a ("clk: bcm: Add BCM2711 DVP driver")
Signed-off-by: Maxime Ripard <maxime@cerno.tech>
Link: https://lore.kernel.org/r/20200626112513.90816-1-maxime@cerno.tech
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

clk: vc5: Enable addition output configurations of the Versaclock

The existing driver is expecting the Versaclock to be pre-programmed,
and only sets the output frequency.  Unfortunately, not all devices
are pre-programmed, and the Versaclock chip has more options beyond
just the frequency.

This patch enables the following additional features:

   - Programmable voltage: 1.8V, 2.5V, or 3.3V​
   - Slew Percentage of normal: 85%, 90%, or 100%
   - Output Type: LVPECL, CMOS, HCSL, or LVDS

Signed-off-by: Adam Ford <aford173@gmail.com>
Link: https://lore.kernel.org/r/20200603154329.31579-3-aford173@gmail.com
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

dt: Add additional option bindings for IDT VersaClock

The VersaClock driver now supports some additional bindings to support
child nodes which can configure optional settings like mode, voltage
and slew.

This patch updates the binding document to describe what is available
in the driver.

Signed-off-by: Adam Ford <aford173@gmail.com>
Reviewed-by: Rob Herring <robh@kernel.org>
Link: https://lore.kernel.org/r/20200603154329.31579-2-aford173@gmail.com
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

clk: vc5: Allow Versaclock driver to support multiple instances

Currently, the Versaclock driver is only expecting one instance and
uses hard-coded names for the various clock names.  Unfortunately,
this is a problem when there is more than one instance of the driver,
because the subsequent instantiations of the driver use the identical
name.  Each clock after the fist fails to load, because the clock
subsystem cannot handle two clocks with identical name.

This patch removes the hard-coded name arrays and uses kasprintf to
assign clock names based on names of their respective node and parent
node which gives each clock a unique identifying name.

For a verasaclock node with a name like:
   versaclock5: versaclock_som@6a

The updated clock names would appear like:
    versaclock_som.mux
       versaclock_som.out0_sel_i2cb
       versaclock_som.pfd
          versaclock_som.pll
             versaclock_som.fod3
                versaclock_som.out4
             versaclock_som.fod2
                versaclock_som.out3
             versaclock_som.fod1
                versaclock_som.out2
             versaclock_som.fod0
                versaclock_som.out1

Signed-off-by: Adam Ford <aford173@gmail.com>
Link: https://lore.kernel.org/r/20200603154329.31579-1-aford173@gmail.com
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

clk: qcom: smd: Add support for MSM8936 rpm clocks

Add missing definition of rpm clk for msm8936 soc (also used by msm8939)

Signed-off-by: Vincent Knecht <vincent.knecht@mailoo.org>
Link: https://lore.kernel.org/r/20200613072745.1249003-2-vincent.knecht@mailoo.org
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

dt-bindings: clock: rpmcc: Document MSM8936 compatible

Acked-by: Rob Herring <robh@kernel.org>
Signed-off-by: Vincent Knecht <vincent.knecht@mailoo.org>
Link: https://lore.kernel.org/r/20200613072745.1249003-3-vincent.knecht@mailoo.org
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

clk: qcom: smd: Add support for SDM660 rpm clocks

Add rpm smd clocks, PMIC and bus clocks which are required on
SDM630/660 (and APQ variants) for clients to vote on.

Signed-off-by: Konrad Dybcio <konradybcio@gmail.com>
Link: https://lore.kernel.org/r/20200622090252.36568-1-konradybcio@gmail.com
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

clk: add function documentation for clk_hw_round_rate()

Information about usage and prerequisites for this API.

Signed-off-by: Sarang Mairal <sarangmairal@gmail.com>
Link: https://lore.kernel.org/r/20200611021941.786-2-sarangmairal@gmail.com
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

clk: qcom: Add ipq6018 apss clock controller

The CPU on Qualcomm ipq6018 devices are clocked primarily by a aplha PLL
and xo which are connected to a mux and enable block.

Add support for the mux and enable block which feeds the CPU on ipq6018
devices.

Reviewed-by: Stephen Boyd <sboyd@kernel.org>
Signed-off-by: Sivaprakash Murugesan <sivaprak@codeaurora.org>
Link: https://lore.kernel.org/r/1592800092-20533-5-git-send-email-sivaprak@codeaurora.org
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

clk: qcom: Add DT bindings for ipq6018 apss clock controller

Add dt-binding for ipq6018 apss clock controller

Acked-by: Rob Herring <robh@kernel.org>
Signed-off-by: Sivaprakash Murugesan <sivaprak@codeaurora.org>
Link: https://lore.kernel.org/r/1592800092-20533-4-git-send-email-sivaprak@codeaurora.org
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

clk: qcom: Add ipq apss pll driver

The CPUs on Qualcomm ipq based devices are clocked by an alpha PLL.
Add support for the apss pll found on ipq based devices which can
support CPU frequencies above 1Ghz.

Signed-off-by: Sivaprakash Murugesan <sivaprak@codeaurora.org>
Link: https://lore.kernel.org/r/1592800092-20533-3-git-send-email-sivaprak@codeaurora.org
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

dt-bindings: clock: add ipq6018 a53 pll compatible

cpus on ipq6018 are clocked by a53 pll, add device compatible for a53
pll found on ipq6018 devices.

Reviewed-by: Rob Herring <robh@kernel.org>
Signed-off-by: Sivaprakash Murugesan <sivaprak@codeaurora.org>
Link: https://lore.kernel.org/r/1592800092-20533-2-git-send-email-sivaprak@codeaurora.org
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

clk: socfpga: agilex: mpu_l2ram_clk should be mpu_ccu_clk

Preliminary documentation documented the mpu_l2ram_clk, but since then,
the mpu_l2ram_clk is no longer documented. It's now referred to as
mpu_ccu_clk.

Signed-off-by: Dinh Nguyen <dinguyen@kernel.org>
Link: https://lore.kernel.org/r/20200616202417.14376-3-dinguyen@kernel.org
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

clk: socfpga: agilex: add nand_x_clk and nand_ecc_clk

And the nand_x_clk and nand_ecc_clk. Make the nand_x_clk be the main
clock that is feeding the NAND IP and correct it's parent to be the
l4_mp_clk.

Signed-off-by: Dinh Nguyen <dinguyen@kernel.org>
Link: https://lore.kernel.org/r/20200616202417.14376-2-dinguyen@kernel.org
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

dt-bindings: agilex: add NAND_X_CLK and NAND_ECC_CLK

Add the NAND_X_CLK and NAND_ECC_CLK clocks.

Signed-off-by: Dinh Nguyen <dinguyen@kernel.org>
Link: https://lore.kernel.org/r/20200616202417.14376-1-dinguyen@kernel.org
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

clk: bcm: rpi: Remove the quirks for the CPU clock

The CPU clock has had so far a bunch of quirks to expose the clock tree
properly, but since we reverted to exposing them through the MMIO driver,
we can remove that code from the firmware driver.

Acked-by: Nicolas Saenz Julienne <nsaenzjulienne@suse.de>
Tested-by: Nicolas Saenz Julienne <nsaenzjulienne@suse.de>
Signed-off-by: Maxime Ripard <maxime@cerno.tech>
Link: https://lore.kernel.org/r/acdf820c2f78a25dd7480a0c018b8b387acd013e.1592210452.git-series.maxime@cerno.tech
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

clk: bcm2835: Don't cache the PLLB rate

The PLLB rate will be changed through the firmware clocks drivers and will
change behind this drivers' back, so we don't want to cache the rate.

Acked-by: Nicolas Saenz Julienne <nsaenzjulienne@suse.de>
Tested-by: Nicolas Saenz Julienne <nsaenzjulienne@suse.de>
Signed-off-by: Maxime Ripard <maxime@cerno.tech>
Link: https://lore.kernel.org/r/9864daba2f584ed49aee5ed1d2f4d48507c58197.1592210452.git-series.maxime@cerno.tech
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

clk: bcm2835: Allow custom CCF flags for the PLLs

While some clock types allow for each clock to specify its own custom
flags, the PLLs can't. We will need this for the PLLB, so let's add it.

Acked-by: Nicolas Saenz Julienne <nsaenzjulienne@suse.de>
Tested-by: Nicolas Saenz Julienne <nsaenzjulienne@suse.de>
Signed-off-by: Maxime Ripard <maxime@cerno.tech>
Link: https://lore.kernel.org/r/ae8bd505d8851f6646e244cd76b6b289346973c8.1592210452.git-series.maxime@cerno.tech
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

Revert "clk: bcm2835: remove pllb"

This reverts commit 2256d89333bd17b8b56b42734a7e1046d52f7fc3. Since we
will be expanding the firmware clock driver, we'll need to remove the
quirks to deal with the PLLB. However, we still want to expose the clock
tree properly, so having that clock in the MMIO driver will allow that.

Acked-by: Nicolas Saenz Julienne <nsaenzjulienne@suse.de>
Tested-by: Nicolas Saenz Julienne <nsaenzjulienne@suse.de>
Signed-off-by: Maxime Ripard <maxime@cerno.tech>
Link: https://lore.kernel.org/r/5d26a4c58248f5be7760a7f2f720a1310baea5dd.1592210452.git-series.maxime@cerno.tech
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

clk: bcm: rpi: Give firmware clocks a name

We've registered the firmware clocks using their ID as name, but it's much
more convenient to register them using their proper name. Since the
firmware doesn't provide it, we have to duplicate it.

Acked-by: Nicolas Saenz Julienne <nsaenzjulienne@suse.de>
Tested-by: Nicolas Saenz Julienne <nsaenzjulienne@suse.de>
Signed-off-by: Maxime Ripard <maxime@cerno.tech>
Link: https://lore.kernel.org/r/a52a5f5768cd33716cdd35237c6613f26ad75013.1592210452.git-series.maxime@cerno.tech
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

clk: bcm: rpi: Discover the firmware clocks

The RaspberryPi4 firmware actually exposes more clocks than are currently
handled by the driver and we will need to change some of them directly
based on the pixel rate for the display related clocks, or the load for the
GPU.

Since the firmware implements DVFS, this rate change can have a number of
side-effects, including adjusting the various PLL voltages or the PLL
parents. The firmware also implements thermal throttling, so even some
thermal pressure can change those parameters behind Linux back.

DVFS is currently implemented on the arm, core, h264, v3d, isp and hevc
clocks, so updating any of them using the MMIO driver (and thus behind the
firmware's back) can lead to troubles, the arm clock obviously being the
most problematic.

In order to make Linux play as nice as possible with those constraints, it
makes sense to rely on the firmware clocks as much as possible. However,
the firmware doesn't seem to provide some equivalents to their MMIO
counterparts, so we can't really replace that driver entirely.

Fortunately, the firmware has an interface to discover the clocks it
exposes.

Let's use it to discover, register the clocks in the clocks framework and
then expose them through the device tree for consumers to use them.

Cc: Michael Turquette <mturquette@baylibre.com>
Cc: Stephen Boyd <sboyd@kernel.org>
Cc: linux-clk@vger.kernel.org
Acked-by: Nicolas Saenz Julienne <nsaenzjulienne@suse.de>
Reviewed-by: Stephen Boyd <sboyd@kernel.org>
Tested-by: Nicolas Saenz Julienne <nsaenzjulienne@suse.de>
Signed-off-by: Maxime Ripard <maxime@cerno.tech>
Link: https://lore.kernel.org/r/438d73962741a8c5f7c689319b7443b930a87fde.1592210452.git-series.maxime@cerno.tech
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

clk: bcm: rpi: Add an enum for the firmware clocks

While the firmware allows us to discover the available clocks, we need to
discriminate those clocks to only register the ones meaningful to Linux.
The firmware also doesn't provide a clock name, so having a list of the ID
will help us to give clocks a proper name later on.

Acked-by: Nicolas Saenz Julienne <nsaenzjulienne@suse.de>
Tested-by: Nicolas Saenz Julienne <nsaenzjulienne@suse.de>
Signed-off-by: Maxime Ripard <maxime@cerno.tech>
Link: https://lore.kernel.org/r/4738f77ee7de9b48a3bb1c558ead958d0cc064d9.1592210452.git-series.maxime@cerno.tech
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

clk: bcm: rpi: Add DT provider for the clocks

For the upcoming registration of the clocks provided by the firmware, make
sure it's exposed to the device tree providers.

Cc: Michael Turquette <mturquette@baylibre.com>
Cc: linux-clk@vger.kernel.org
Acked-by: Nicolas Saenz Julienne <nsaenzjulienne@suse.de>
Reviewed-by: Stephen Boyd <sboyd@kernel.org>
Tested-by: Nicolas Saenz Julienne <nsaenzjulienne@suse.de>
Signed-off-by: Maxime Ripard <maxime@cerno.tech>
Link: https://lore.kernel.org/r/4d8dbe4aaae98b3d3812ad7c3dba53d645cadbaf.1592210452.git-series.maxime@cerno.tech
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

clk: bcm: rpi: Make the PLLB registration function return a clk_hw

The raspberrypi_register_pllb has been returning an integer so far to
notify whether the functions has exited successfully or not.

However, the OF provider functions in the clock framework require access to
the clk_hw structure so that we can expose those clocks to device tree
consumers.

Since we'll want that for the future clocks, let's return a clk_hw pointer
instead of the return code.

Cc: Michael Turquette <mturquette@baylibre.com>
Cc: linux-clk@vger.kernel.org
Acked-by: Nicolas Saenz Julienne <nsaenzjulienne@suse.de>
Reviewed-by: Stephen Boyd <sboyd@kernel.org>
Tested-by: Nicolas Saenz Julienne <nsaenzjulienne@suse.de>
Signed-off-by: Maxime Ripard <maxime@cerno.tech>
Link: https://lore.kernel.org/r/97218559db643e62fdd2b5e3046a2a05b8c2e769.1592210452.git-series.maxime@cerno.tech
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

clk: bcm: rpi: Split pllb clock hooks

The driver only supports the pllb for now and all the clock framework hooks
are a mix of the generic firmware interface and the specifics of the pllb.
Since we will support more clocks in the future let's split the generic and
specific hooks

Cc: Michael Turquette <mturquette@baylibre.com>
Cc: linux-clk@vger.kernel.org
Acked-by: Nicolas Saenz Julienne <nsaenzjulienne@suse.de>
Reviewed-by: Stephen Boyd <sboyd@kernel.org>
Tested-by: Nicolas Saenz Julienne <nsaenzjulienne@suse.de>
Signed-off-by: Maxime Ripard <maxime@cerno.tech>
Link: https://lore.kernel.org/r/fdc21962fdc7de5c46232f198672d5d5c868ec74.1592210452.git-series.maxime@cerno.tech
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

clk: bcm: rpi: Rename is_prepared function

The raspberrypi_fw_pll_is_on function doesn't only apply to PLL
registered in the driver, but any clock exposed by the firmware.

Since we also implement the is_prepared hook, make the function
consistent with the other function names.

Cc: Michael Turquette <mturquette@baylibre.com>
Cc: linux-clk@vger.kernel.org
Acked-by: Nicolas Saenz Julienne <nsaenzjulienne@suse.de>
Reviewed-by: Stephen Boyd <sboyd@kernel.org>
Tested-by: Nicolas Saenz Julienne <nsaenzjulienne@suse.de>
Signed-off-by: Maxime Ripard <maxime@cerno.tech>
Link: https://lore.kernel.org/r/ac93cc4e245316bb7e7426ac5ab0de8f3d919731.1592210452.git-series.maxime@cerno.tech
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

clk: bcm: rpi: Pass the clocks data to the firmware function

The raspberry_clock_property only takes the clock ID as an argument, but
now that we have a clock data structure it makes more sense to just pass
that structure instead.

Cc: Michael Turquette <mturquette@baylibre.com>
Cc: Stephen Boyd <sboyd@kernel.org>
Cc: linux-clk@vger.kernel.org
Acked-by: Nicolas Saenz Julienne <nsaenzjulienne@suse.de>
Reviewed-by: Stephen Boyd <sboyd@kernel.org>
Tested-by: Nicolas Saenz Julienne <nsaenzjulienne@suse.de>
Signed-off-by: Maxime Ripard <maxime@cerno.tech>
Link: https://lore.kernel.org/r/d7a3b4df3ca23feb6e0d9c7ae2d232bfb913f926.1592210452.git-series.maxime@cerno.tech
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

clk: bcm: rpi: Add clock id to data

The driver has really only supported one clock so far and has hardcoded the
ID used in communications with the firmware in all the functions
implementing the clock framework hooks. Let's store that in the clock data
structure so that we can support more clocks later on.

Cc: Michael Turquette <mturquette@baylibre.com>
Cc: linux-clk@vger.kernel.org
Acked-by: Nicolas Saenz Julienne <nsaenzjulienne@suse.de>
Reviewed-by: Stephen Boyd <sboyd@kernel.org>
Tested-by: Nicolas Saenz Julienne <nsaenzjulienne@suse.de>
Signed-off-by: Maxime Ripard <maxime@cerno.tech>
Link: https://lore.kernel.org/r/e23c37961b97b027e21efa3b818578970f88527a.1592210452.git-series.maxime@cerno.tech
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

clk: bcm: rpi: Create a data structure for the clocks

So far the driver has really only been providing a single clock, and stored
both the data associated to that clock in particular with the data
associated to the "controller".

Since we will change that in the future, let's decouple the clock data from
the provider data.

Cc: Michael Turquette <mturquette@baylibre.com>
Cc: linux-clk@vger.kernel.org
Acked-by: Nicolas Saenz Julienne <nsaenzjulienne@suse.de>
Reviewed-by: Stephen Boyd <sboyd@kernel.org>
Tested-by: Nicolas Saenz Julienne <nsaenzjulienne@suse.de>
Signed-off-by: Maxime Ripard <maxime@cerno.tech>
Link: https://lore.kernel.org/r/ee7f508db226214fab4add7f93a351f4137c86a1.1592210452.git-series.maxime@cerno.tech
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

clk: bcm: rpi: Use CCF boundaries instead of rolling our own

The raspberrypi firmware clock driver has a min_rate / max_rate clamping by
storing the info it needs in a private structure.

However, the CCF already provides such a facility, so we can switch to it
to remove the boilerplate.

Reviewed-by: Nicolas Saenz Julienne <nsaenzjulienne@suse.de>
Tested-by: Nicolas Saenz Julienne <nsaenzjulienne@suse.de>
Signed-off-by: Maxime Ripard <maxime@cerno.tech>
Link: https://lore.kernel.org/r/d4c53dab6de5d5f70743d9c139d0117589530e62.1592210452.git-series.maxime@cerno.tech
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

clk: bcm: rpi: Make sure the clkdev lookup is removed

The clkdev lookup created for the cpufreq device is never removed if
there's an issue later in probe or at module removal time.

Let's convert to the managed variant of the clk_hw_register_clkdev function
to make sure it happens.

Cc: Michael Turquette <mturquette@baylibre.com>
Cc: linux-clk@vger.kernel.org
Acked-by: Nicolas Saenz Julienne <nsaenzjulienne@suse.de>
Reviewed-by: Stephen Boyd <sboyd@kernel.org>
Tested-by: Nicolas Saenz Julienne <nsaenzjulienne@suse.de>
Signed-off-by: Maxime Ripard <maxime@cerno.tech>
Link: https://lore.kernel.org/r/075e2c6d315eccdaf8fb72b320712b86e6c25b22.1592210452.git-series.maxime@cerno.tech
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

clk: bcm: rpi: Switch to clk_hw_register_clkdev

Since we don't care about retrieving the clk_lookup structure pointer
returned by clkdev_hw_create, we can just use the clk_hw_register_clkdev
function.

Cc: Michael Turquette <mturquette@baylibre.com>
Cc: linux-clk@vger.kernel.org
Acked-by: Nicolas Saenz Julienne <nsaenzjulienne@suse.de>
Reviewed-by: Stephen Boyd <sboyd@kernel.org>
Tested-by: Nicolas Saenz Julienne <nsaenzjulienne@suse.de>
Signed-off-by: Maxime Ripard <maxime@cerno.tech>
Link: https://lore.kernel.org/r/59f6208b6fe3367e735b0cca4f65c2c937639af9.1592210452.git-series.maxime@cerno.tech
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

clk: bcm: rpi: Remove pllb_arm_lookup global pointer

The pllb_arm_lookup pointer in the struct raspberrypi_clk is not used for
anything but to store the returned pointer to clkdev_hw_create, and is not
used anywhere else in the driver.

Let's remove that global pointer from the structure.

Cc: Michael Turquette <mturquette@baylibre.com>
Cc: linux-clk@vger.kernel.org
Acked-by: Nicolas Saenz Julienne <nsaenzjulienne@suse.de>
Reviewed-by: Stephen Boyd <sboyd@kernel.org>
Tested-by: Nicolas Saenz Julienne <nsaenzjulienne@suse.de>
Signed-off-by: Maxime Ripard <maxime@cerno.tech>
Link: https://lore.kernel.org/r/189407f54906d2b07c91de7a4eeb6d8c8934280f.1592210452.git-series.maxime@cerno.tech
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

clk: bcm: rpi: Make sure pllb_arm is removed

The pllb_arm clock was created at probe time, but was never removed if
something went wrong later in probe, or if the driver was ever removed from
the system.

Now that we are using clk_hw_register(), we can just use its managed variant
to take care of that for us.

Cc: Michael Turquette <mturquette@baylibre.com>
Cc: Stephen Boyd <sboyd@kernel.org>
Cc: linux-clk@vger.kernel.org
Acked-by: Nicolas Saenz Julienne <nsaenzjulienne@suse.de>
Reviewed-by: Stephen Boyd <sboyd@kernel.org>
Tested-by: Nicolas Saenz Julienne <nsaenzjulienne@suse.de>
Signed-off-by: Maxime Ripard <maxime@cerno.tech>
Link: https://lore.kernel.org/r/34254ed1556614658e5dad5cca4cf4fe617df7fc.1592210452.git-series.maxime@cerno.tech
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

clk: bcm: rpi: Remove global pllb_arm clock pointer

The pllb_arm clk_hw pointer in the raspberry_clk structure isn't used
anywhere but in the raspberrypi_register_pllb_arm.

Let's remove it, this will make our lives easier in future patches.

Cc: Michael Turquette <mturquette@baylibre.com>
Cc: Stephen Boyd <sboyd@kernel.org>
Cc: linux-clk@vger.kernel.org
Acked-by: Nicolas Saenz Julienne <nsaenzjulienne@suse.de>
Reviewed-by: Stephen Boyd <sboyd@kernel.org>
Tested-by: Nicolas Saenz Julienne <nsaenzjulienne@suse.de>
Signed-off-by: Maxime Ripard <maxime@cerno.tech>
Link: https://lore.kernel.org/r/842859cf1a77478620f45049178a588448202858.1592210452.git-series.maxime@cerno.tech
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

clk: bcm: rpi: Use clk_hw_register for pllb_arm

The pllb_arm clock is defined as a fixed factor clock with the pllb
clock as a parent. However, all its configuration is entirely static,
and thus we don't really need to call clk_hw_register_fixed_factor() but
can simply call clk_hw_register() with a static clk_fixed_factor
structure.

Cc: Michael Turquette <mturquette@baylibre.com>
Cc: linux-clk@vger.kernel.org
Acked-by: Nicolas Saenz Julienne <nsaenzjulienne@suse.de>
Reviewed-by: Stephen Boyd <sboyd@kernel.org>
Tested-by: Nicolas Saenz Julienne <nsaenzjulienne@suse.de>
Signed-off-by: Maxime Ripard <maxime@cerno.tech>
Link: https://lore.kernel.org/r/1146177664999eeda65856d28ce94025021dd85e.1592210452.git-series.maxime@cerno.tech
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

clk: bcm: rpi: Statically init clk_init_data

Instead of declaring the clk_init_data and then calling memset on it, just
initialise properly.

Cc: Michael Turquette <mturquette@baylibre.com>
Cc: Stephen Boyd <sboyd@kernel.org>
Cc: linux-clk@vger.kernel.org
Acked-by: Nicolas Saenz Julienne <nsaenzjulienne@suse.de>
Reviewed-by: Stephen Boyd <sboyd@kernel.org>
Tested-by: Nicolas Saenz Julienne <nsaenzjulienne@suse.de>
Signed-off-by: Maxime Ripard <maxime@cerno.tech>
Link: https://lore.kernel.org/r/0342572daa561dc1bb4c9fd10641b2016493e32b.1592210452.git-series.maxime@cerno.tech
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

clk: bcm: rpi: Allow the driver to be probed by DT

The current firmware clock driver for the RaspberryPi can only be probed by
manually registering an associated platform_device.

While this works fine for cpufreq where the device gets attached a clkdev
lookup, it would be tedious to maintain a table of all the devices using
one of the clocks exposed by the firmware.

Since the DT on the other hand is the perfect place to store those
associations, make the firmware clocks driver probe-able through the device
tree so that we can represent it as a node.

Cc: Michael Turquette <mturquette@baylibre.com>
Cc: linux-clk@vger.kernel.org
Reviewed-by: Nicolas Saenz Julienne <nsaenzjulienne@suse.de>
Reviewed-by: Stephen Boyd <sboyd@kernel.org>
Tested-by: Nicolas Saenz Julienne <nsaenzjulienne@suse.de>
Signed-off-by: Maxime Ripard <maxime@cerno.tech>
Link: https://lore.kernel.org/r/cb8203b862e386ac6c3df3eff0bb5a238b6ec97a.1592210452.git-series.maxime@cerno.tech
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

firmware: rpi: Only create clocks device if we don't have a node for it

The firmware clocks driver was previously probed through a platform_device
created by the firmware driver.

Since we will now have a node for that clocks driver, we need to create the
device only in the case where there's no node for it already.

Reviewed-by: Nicolas Saenz Julienne <nsaenzjulienne@suse.de>
Tested-by: Nicolas Saenz Julienne <nsaenzjulienne@suse.de>
Signed-off-by: Maxime Ripard <maxime@cerno.tech>
Link: https://lore.kernel.org/r/72114c4287ebda2dbd952ea238d4489d359897e5.1592210452.git-series.maxime@cerno.tech
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

dt-bindings: clock: Add a binding for the RPi Firmware clocks

The firmware running on the RPi VideoCore can be used to discover and
change the various clocks running in the BCM2711. Since devices will
need to use them through the DT, let's add a pretty simple binding.

Cc: Michael Turquette <mturquette@baylibre.com>
Cc: linux-clk@vger.kernel.org
Cc: devicetree@vger.kernel.org
Reviewed-by: Stephen Boyd <sboyd@kernel.org>
Reviewed-by: Rob Herring <robh+dt@kernel.org>
Tested-by: Nicolas Saenz Julienne <nsaenzjulienne@suse.de>
Signed-off-by: Maxime Ripard <maxime@cerno.tech>
Link: https://lore.kernel.org/r/f6105207e7ef5a5ea8d7a1774faf989d341a25f5.1592210452.git-series.maxime@cerno.tech
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

dt-bindings: arm: bcm: Convert BCM2835 firmware binding to YAML

Convert the Raspberry Pi BCM2835 firmware binding document to YAML.
Verified with dt_binding_check and dtbs_check.

Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
Tested-by: Nicolas Saenz Julienne <nsaenzjulienne@suse.de>
Signed-off-by: Maxime Ripard <maxime@cerno.tech>
Link: https://lore.kernel.org/r/5bc0b9be8544b07300fccab4d4f26e5e5d8e62b2.1592210452.git-series.maxime@cerno.tech
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

clk: bcm: Add BCM2711 DVP driver

The HDMI block has a block that controls clocks and reset signals to the
HDMI0 and HDMI1 controllers.

Let's expose that through a clock driver implementing a clock and reset
provider.

Cc: Michael Turquette <mturquette@baylibre.com>
Cc: Stephen Boyd <sboyd@kernel.org>
Cc: Rob Herring <robh+dt@kernel.org>
Cc: linux-clk@vger.kernel.org
Cc: devicetree@vger.kernel.org
Reviewed-by: Stephen Boyd <sboyd@kernel.org>
Signed-off-by: Maxime Ripard <maxime@cerno.tech>
Link: https://lore.kernel.org/r/bb60d97fc76b61c2eabef5a02ebd664c0f57ede0.1591867332.git-series.maxime@cerno.tech
Acked-by: Stefan Wahren <stefan.wahren@i2se.com>
Reviewed-by: Nicolas Saenz Julienne <nsaenzjulienne@suse.de>
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

dt-bindings: clock: Add BCM2711 DVP binding

The BCM2711 has a unit controlling the HDMI0 and HDMI1 clock and reset
signals. Let's add a binding for it.

Cc: Philipp Zabel <p.zabel@pengutronix.de>
Cc: Rob Herring <robh+dt@kernel.org>
Cc: devicetree@vger.kernel.org
Reviewed-by: Rob Herring <robh+dt@kernel.org>
Signed-off-by: Maxime Ripard <maxime@cerno.tech>
Link: https://lore.kernel.org/r/0b8f09baff1ff3c471631e6f523e2b2cd773ec47.1591867332.git-series.maxime@cerno.tech
Acked-by: Stefan Wahren <stefan.wahren@i2se.com>
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

Merge branch 'reset/simple' of git://git.pengutronix.de/git/pza/linux into clk-bcm

Pull immutable reset branch to get reset-simple header.

* 'reset/simple' of git://git.pengutronix.de/git/pza/linux:
  reset: simple: Add reset callback
  reset: Move reset-simple header out of drivers/reset

clk: bcm63xx-gate: switch to dt-bindings definitions

Now that there are header files for each SoC, let's use them in the
bcm63xx-gate controller driver.

Signed-off-by: Álvaro Fernández Rojas <noltari@gmail.com>
Link: https://lore.kernel.org/r/20200615090231.2932696-9-noltari@gmail.com
Acked-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

mips: bmips: add BCM63268 clock definitions

Add header with BCM63268 definitions in order to be able to include it from
device tree files.

Signed-off-by: Álvaro Fernández Rojas <noltari@gmail.com>
Link: https://lore.kernel.org/r/20200615090231.2932696-8-noltari@gmail.com
Acked-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

mips: bmips: add BCM6368 clock definitions

Add header with BCM6368 definitions in order to be able to include it from
device tree files.

Signed-off-by: Álvaro Fernández Rojas <noltari@gmail.com>
Link: https://lore.kernel.org/r/20200615090231.2932696-7-noltari@gmail.com
Acked-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

mips: bmips: add BCM6362 clock definitions

Add header with BCM6362 definitions in order to be able to include it from
device tree files.

Signed-off-by: Álvaro Fernández Rojas <noltari@gmail.com>
Link: https://lore.kernel.org/r/20200615090231.2932696-6-noltari@gmail.com
Acked-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

mips: bmips: add BCM6358 clock definitions

Add header with BCM6358 definitions in order to be able to include it from
device tree files.

Signed-off-by: Álvaro Fernández Rojas <noltari@gmail.com>
Link: https://lore.kernel.org/r/20200615090231.2932696-5-noltari@gmail.com
Acked-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

mips: bmips: add BCM6328 clock definitions

Add header with BCM6328 definitions in order to be able to include it from
device tree files.

Signed-off-by: Álvaro Fernández Rojas <noltari@gmail.com>
Link: https://lore.kernel.org/r/20200615090231.2932696-4-noltari@gmail.com
Acked-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

mips: bmips: add BCM6318 clock definitions

Add header with BCM6318 definitions in order to be able to include it from
device tree files.

Signed-off-by: Álvaro Fernández Rojas <noltari@gmail.com>
Link: https://lore.kernel.org/r/20200615090231.2932696-3-noltari@gmail.com
Acked-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

mips: bmips: add BCM3368 clock definitions

Add header with BCM3368 definitions in order to be able to include it from
device tree files.

Signed-off-by: Álvaro Fernández Rojas <noltari@gmail.com>
Link: https://lore.kernel.org/r/20200615090231.2932696-2-noltari@gmail.com
Acked-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

clk: bcm63xx-gate: add BCM6318 support

Add support for the gated clock controllers found on the BCM6318.

Signed-off-by: Álvaro Fernández Rojas <noltari@gmail.com>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Link: https://lore.kernel.org/r/20200610140858.207329-3-noltari@gmail.com
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

dt-bindings: clock: bcm63xx: add 6318 gated clock bindings

Add BCM6318 to the binding documentation for the gated clock controllers found
on BCM63xx SoCs.

Signed-off-by: Álvaro Fernández Rojas <noltari@gmail.com>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Link: https://lore.kernel.org/r/20200610140858.207329-2-noltari@gmail.com
Acked-by: Rob Herring <robh@kernel.org>
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

clk: bcm63xx-gate: fix last clock availability

In order to make the last clock available, maxbit has to be set to the
highest bit value plus 1.

Fixes: 1c099779c1e2 ("clk: add BCM63XX gated clock controller driver")
Signed-off-by: Álvaro Fernández Rojas <noltari@gmail.com>
Link: https://lore.kernel.org/r/20200609110846.4029620-1-noltari@gmail.com
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: Stephen Boyd <sboyd@kernel.org>

reset: simple: Add reset callback

The reset-simple code lacks a reset callback that is still pretty easy to
implement. The only real thing to consider is the delay needed for a device
to be reset, so let's expose that as part of the reset-simple driver data.

Reviewed-by: Philipp Zabel <p.zabel@pengutronix.de>
Signed-off-by: Maxime Ripard <maxime@cerno.tech>
Signed-off-by: Philipp Zabel <p.zabel@pengutronix.de>

reset: Move reset-simple header out of drivers/reset

The reset-simple code can be useful for drivers outside of drivers/reset
that have a few reset controls as part of their features. Let's move it to
include/linux/reset.

Reviewed-by: Philipp Zabel <p.zabel@pengutronix.de>
Signed-off-by: Maxime Ripard <maxime@cerno.tech>
Signed-off-by: Philipp Zabel <p.zabel@pengutronix.de>

Linux 5.8-rc1